Monday, September 30, 2019

Trade Desk (TTD) Can Double: TAM Analysis and Valuation

I think the Trade Desk (TTD) can grow its revenue 5x in 5 years, and the stock can double.

Trade Desk is a Demand Side Platform (“DSP”) in the advertising world. The easiest way to describe TTD is via analogy. When you want to buy stocks you would log into a platform like Interactive Brokers, Fidelity or Schwab to put in your buy orders. These platforms would execute your orders by going to the stock exchanges (directly or indirectly), and you can keep track of your portfolio.

In the advertising world, agencies work on behalf of brands to buy ads from publishers. These agencies would use the Trade Desk’s platform to put in buy orders.

Here’s a simplified lay of the land:

Source here.

In the above diagram, Trade Desk takes the role of “DSP”, and transact in the market place. "Buy” and “sell” here refer to advertising inventories, so that anything from DSP and to the left is considered “buy side”, while anything from SSP and to the right is the “sell side”.

A Bull Market in Connected TV ("CTV")

The entire digital advertising space is undergoing expansion. Advertising has been steadily going digital the past decade, but a large budget for traditional TV ads remained.

That is being chipped away. Content that used to go on traditional TVs are all going online, first with Netflix, then HBO, Disney...etc. Smart TV boxes like Roku and Amazon Fire TV are accelerating that trend by aggregating content and making the user experience more convenient.

The most premium of these services like Netflix will be subscription based. But there is a long tail of contents that needs to go online, and many (if not most) of them will be funded with advertisements.

Ad buyers are increasingly seeing the benefit of CTV advertising - more precise ad placements, better customer experiences (frequency capping), better measures, more flexibility, and so on.

Roku’s streaming hours grew 72% yoy and its platform revenue grew 86% in the latest quarter. Telaria’s CTV business grew 133% yoy. Trade Desk saw CTV spend grew 2.5x yoy.

Now, that’s what I call a bull market!

All three companies mentioned above are buys in my opinion, with varying degrees of risks and rewards. I choose to focus on TTD here because I think it has the most visible future.

Roku is in land grab mode against Google and Amazon to embed its platform into smart TVs. Victory is far from assured given competitors' advantage in voice control technology. Telaria is a sell side platform, which is to say there’s about zero barrier to entry, since there’s always new publishers popping up. Also, a publisher may develop its own sell-side ad tech as it gets big (Opera is one such example). Ad revenues are central to the business of content owners after all.

The DSP's do not have this problem. Advertising spend is important to brands, but not so central to the business that brands would develop its own DSP – at least I have not heard of any such examples. In any case, brands mostly outsource to advertising agencies. There’s only a few of these (Omnicom, WPP, IPG, Publicis, Dentsu), and they have been using the Trade Desk.

Although I focus on CTV above, TTD will also benefit from other factors, including programmatic ad buying taking more share within digital advertising.

There are some moats here. There’s cross-side network effect where ad buyers want platforms that can access the most sell side inventories, and vice versa. In addition, there are only so many platforms that ad buyers would use, maybe one to three, and that’s it. (Go back to the stock buying example, how many brokerage platforms do you use?) It won’t be totally an oligopoly as switching costs and barriers to entry are not insurmountable, but industry structure should be fairly stable and concentrated in a few strong hands.

Trade Desk’s Market Position Within DSPs

Most agencies use 2-3 DSPs. As the “Usage” table below shows, it is basically platforms from Amazon, Google and TTD. Note that even though FB takes a lot of ad dollars they're not a DSP player.

Source: TTD Investor Day

TTD’s position looks even stronger if you consider that Amazon actually let Trade Desk bid on its CTV advertising inventories.

The other credible competitors are Adobe and MediaMath. I consider Adobe to be the bigger threat just from a sheer resource perspective, and I would note that Adobe is an “independent” that doesn’t own content/ad inventory, unlike Google or Amazon. This lack of conflict of interest gave TTD an edge over other platforms, and it might also give Adobe an edge.

We will have to factor in Adobe in the market share assumptions below.

Total Addressable Market analysis (U.S. only)

Note this analysis is U.S. only. I was going to do global but it turns out even just U.S. TAM supports a doubling of TTD’s share price. So I’m content using just U.S. TAM/revenues in my valuation while knowing there’s a huge international upside that I’m not factoring in. 

First, some facts:
  • Total U.S. ad spend $207bn in 2018; of which ~$106bn are digital ad spends and ~$60bn are TV ads. (See table below).
  • About 2/3 of digital ads go through Google and Facebook (“walled gardens”)
  • Only about $60 billion of ad spends are bought programmatically, but that number is growing fast.

Here’s a core assumption in my TAM estimate: eventually close to 100% of ad buying will be digital and programmatic (including programmatic direct which TTD supports). It just makes sense. I have not seen any evidence that this is not a safe bet.

I will also estimate TV and non-TV separately, since the former does not have the “walled garden” domination of Facebook and Google, and thus leaving more room for Trade Desk. Amazon’s opening up its CTV inventories is a strong signal that open internet will be the paradigm here.

So here’s my analysis of TAM/market share/and Trade Desk’s peak revenue, all in 2018 dollars.

1a): U.S. non-TV (~$107bn)
  • The big wall gardens like Google and Facebook own about 65-75% of this market. That leaves 25-35% of TAM for the open internet players, or ~$32bn.
  • DSP take rates of 15% (current take rate about 20% but I assume it will degrade over the years).
  • $4.8bn ($32bn* 15%) of potential revenues for Trade Desk and other DSPs to fight over.
1b) U.S. TV (~$60bn)
  • Again, I believe eventually all the ads will be digital and programmatic. No “walled garden” here.
  • DSP take rate of 10-15%. I assume a lower take rate here than non-TV because there are multiple layers of distribution and aggregation like Roku and Amazon Fire TV taking their cut already.
  • $6bn ($60 * 10%) of potential revenues for Trade Desk and other DSPs

2) TTD's market share and U.S. revenue?
  • Non-TV
    • 50/50 split between TTD and Adobe (remember this is assuming the walled gardens like Google and Facebook already took their share)
    • TTD's revenue = $2.4bn (half of the $4.8bn revenue for DSPs)
  • TV
    • Split between 5 DSP players: Amazon, Google, TTD, Adobe, and some others
    • TTD's revenue = $1.2bn ($6bn / 5)

Add up Non-TV and TV, I project TTD can eventually get to $3.6bn in revenue. The company will likely do ~$650mm of revenues for 2019. So at the current 40% revenue growth rate, TTD can get to this $3.6bn in 5 years.

There are multiple elements of conservatism in the above estimate:
  • I'm using 2018 dollars. 
  • Remember, that’s just the U.S. TTD is making a big effort to grow internationally, particularly in China, and I believe they will be successful.
  • I’m assuming TV ad market size stays the same even with various new innovations. The more realistic assumption would be to assume TAM expansion as technology drives down cost and improve access and adoption.
  • I'm not counting audio at all. This is a market that’s going through hyper-growth just like the CTV market.
  • The Walled Garden in non-TV advertising could come down due to regulatory pressures on Facebook and Google. This would instantly triple TTD’s opportunity in that area.

Stock Valuation and Upside

Take the $3.6bn revenue from above and apply 30% EBITDA margin and 20x EV/EBITDA (still plenty of growth runway in 5 years, strong market position, software margins…etc) get you ~21.5bn of enterprise value. Assume a little share dilution and stock is a double in 5 years for 15% IRR.


Take rates can come down to even lower than I modeled.

No comments:

Post a Comment